International Conference on Civil Engineering

Architecture & Urban Sustainable Development

18 &19 December 2013, Tabriz , Iran

Evaluation of Site Effects on Strong Ground Motion

Abdolah Tabaroei*1

1. Department of Civil, Science and Research Branch, Islamic Azad University, Tehran, Iran,
tabaroei_a@srbiau.ac.ir

Abstract
Nowadays, earthquake is one of the most devastating natural disasters that can be
somewhat resisted by reinforcing structures. From an engineering perspective, the
importance of an earthquake is due to it’s effect on structures such as residential
housings, dams, bridges, nuclear facilities, industrial installations and so on. Evaluating
the ground response is one of the main issues in geotechnical engineering which has
raised new challenges for engineers. Ground response analysis is used for predicting
ground surface motions. Among the factors influencing the prediction of strong ground
motion is site effects. These effects include source, path and site effects, topography
effects and soil layering effects. In this paper, the results of accelerographs are first
modified by using the SEISMOSIGNAL software (band-path filter of 0.2-30 Hz and
modifying the baseline), and then by using DEEPSOIL software, the results are analyzed
directly from station (1) to station (2) and by deconvolution from station (2) to station (1)
in both time and frequency domain.

Key words: Site Effect, Strong Ground Motion, SEISMOSIGNAL, DEEPSOIL Software, Time
Domain, Frequency Domain

1. Introduction
From an engineering point of view, the significance of earthquakes is due to it’s effects
made on structures such as dams (concrete or earth), powerhouses, bridges, embankments,
residential housings, and industrial installations, which in most cases these structures are built
on alluvial layers placed on bedrock, not on the surface of rock masses. These structures are
also placed on various distances from hypocenter and mostly, they are not built on the fault
location or the starting location of seismic waves. Receding from the alluvial layers, the
earthquake waves will be exposed to changes which are known as the factors related to the

1
Corresponding Author, Tel: +98-9128464306

1
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

distance of the points from epicenter referred to as “path effect” and the factors related to
alluvial layers placed on the bedrock referred to as “site effect” [1]. As mentioned above, the
issue of local side effects is one of the most important subjects in seismic geotechnical
engineering. Mac Mardu (1824) noted in his report of 1819’s earthquake in Cutch (India) that
the structures placed on stones were not affected by earthquake as much as the structures
whose foundations were placed on the bottom layers of soil. By investigating the ground’s
response during the 1994 earthquake in Nortridge near Los Angeles and comparing the results
with the results of equivalent linear and non-linear analysis, Chang et al. (1996) insisted on
the significance of deep and hard alluvial layers of the area in intensifying ground motions
[2]. Generally, many methods have been proposed for estimating the site responses since
1920. First it was looked at the site effect phenomena from the linear elasticity theory and
reflection theory’s point of view, and then average linear models were widely used for
modeling the effect of nonlinear soil. Joyner & Chen (1957) and Hwang & Lee (1991)
investigated the site parameters such as secant shear modulus, damping ratio at low strains,
various types of soil (clay and sand), underground water location, and bedrock’s depth [3,4].
Since the early 20th century, the subject of site effect on the seismic ground motions was of
interest to researchers and scientists. They evaluated the site effect in changing the various
properties of earthquake waves by investigating the effects caused by earthquake waves in
different places of site and also by investigating the conditions of alluvial layers of the
location. These investigations have been continued up to recent years, and after occurrence of
new earthquakes in different parts of the world, many researches were done about how these
site effects influenced on the earthquake intensity. Local conditions of the site have a great
effect on the total important properties of strong ground motion such as domain, frequency
content, and duration. According to the evidences from past earthquakes in some places, the
site resonance factor has reached 5, while in the USB regulations, the amount of site factor (S)
is considered 1 to 2 and in the Iran’s regulations of 2800 it is considered 0.6 to 2.5 [5]. In this
paper after modifying and filtering the results of the accelerographs, we have used
DEEPSOIL software for direct and deconvolution analysis in time and frequency domain.

2. DEEPSOIL Software
DEEPSOIL is a one-dimensional site response analysis program that can perform both in
one dimensional nonlinear and equivalent linear analyses and features an intuitive graphical
user interface. The program is provided as-is and the user assumes responsibility for all
results. We requires knowledge in the theory and procedures for seismic site response
analysis. It is suggested that the user reviews relevant literature [6].

2.1. DEEPSOIL Structure

2
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

The DEEPSOIL graphical user interface is composed of 5 (for equivalent linear) and 6 (for
nonlinear) stages and intuitively guides the user from the beginning to the end of the site
response analysis. The logic is mapped out in the flowchart below:

Initialization

Analysis Type Selection

Define Soil Profile and Model Properties

Analysis Control

Motion and Output Control

Time Yes Viscous Damping

Domain Formulation Type/
Analysis ? Optimum Modes Selection

No

Output

Figure 1.DEEPSOIL software flowchart [6]

2.2. Analysis Type Selection

3
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

The first step in the analysis requires the selection of analysis type [6]. To create a new
analysis, we must specify the type of analysis before proceeding to the next stage of analysis.
We must specify [6]:
1. The number of layers to be used in the profile.
2. The analysis method:
 Frequency Domain
 Linear
 Equivalent Linear
 Time Domain
 Linear
 Nonlinear
3. The type of input for shear properties:
 Shear Modulus
 Shear Wave Velocity
4. The units to be used in analysis:
 English
 Metric
5. The analysis type:
 Total Stress Analysis
 Effective Stress Analysis (Pore Water Pressure generation only)
 Include PWP Dissipation (PWP generation and dissipation)
6. The Method to define the soil curve:
 For Equivalent Linear
 Discrete Points
 Pressure- Dependent Hyperbolic Model
 For Nonlinear
 MRDF Pressure- Dependent Hyperbolic Model
 Pressure- Dependent Hyperbolic Model
7. The boundary conditions (for Effective Stress Analysis Incl. PWP Dissipation)
The Effective Stress Analysis option is only available for a Nonlinear (The Domain) analysis.
Note that (1), (3), and (4) can be changed in the next stage [6].

2.3. Defining Soil Profile and Model Properties

This stage is divided into two partitions. The first partition to be considered requires we to
define the soil profile and specify the soil properties of each layer. The type of input required
depends on the analysis parameters selected in step 1.

4
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

2.4. Soil Properties

A variety of models are available for DEEPSOIL analyses. These models include:
1) Equivalent Linear
2) Hyperbolic (MR, MRD, DC)
3) A new Hyperbolic model (MRDF)
4) Porewater Pressure Generation and Dissipation
Material properties of soil in the analysis are listed in Table (1).

Table 1: Material Properties of soil

Depth Unit Weight VP VS

Layer No. Soil Type
(m) (gr/cm3) (m/s) (m/s)
1 Surface Layer 0-3 1.7 745 435

2 Silty Sand (Satured) 3-30 1.9 1705 550

3 Silty Clay (Dense) 30-40 2 1800 665

2.5. Define Rock Properties

After defining the soil and model properties, we must now define the rock/ half-space
properties of the bottom of the profile [6]. Material properties of bedrock in the analysis are
listed in Table (2).

Table 1: Material Properties of Bedrock

Depth Unit Weight VP VS

Bedrock No.
(m) (gr/cm3) (m/s) (m/s)

1 >40 2 2000 800

2.5. Define Rock Properties

5
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

In this stage of analysis, we may specify specific options to be used for either the frequency
domain or time domain analysis [6].

2.6. Motion and Output Control

Depending on the user’s selected option during the Initialization stage, the options available
in this stage of analysis will vary depending on if a standard or batch mode analysis is being
performed. In both cases, the input motion (s) and layer (s) for output display will be selected.
We should also choose the damping ratio for the calculated response spectra. The motion
control stage allows the user to specify the input motion to be used in analysis and selection
of the layers to be analyzed. The number of calculation points is only relevant in the
frequency domain and should be specified when using the Fast Fourier Transform [6].

2.7. Output

Upon completion of analysis, the following output for each selected layer will be directly
exported to a text file “ExportOutpu.txt” in the working directory [6].

3. The Result of Analysis

3.1. Direct Analysis in Time Domain

As we mentioned earlier, the DEEPSOIL software is used for direct and deconvolution
analysis, in which the results of time history of given acceleration is considered as input
motion at the surface of outcrop in station (1) for direct analysis and the output in station (2)
placed on the ground surface is obtained from the software, and then the results obtained from
the software is compared with the results obtained from accelerographs in time domain in
order to ensure the accuracy of the results. All the steps above are done for east-west, north-
south, and vertical components and the results are as follows.

6
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 2: A view of direct analysis

The East-west Component of Earthquake

Figure 3: The time history of acceleration for east-west component obtained in station 2 (In-situ
measurements and calculated values)

As it’s demonstrated in figure (1), The east-west input motion of station (1) in the surface of
outcrop is given to the software and the output motion of station (2) in the ground surface is
obtained from the software or in the other words, the direct analysis has taken place. In order
to compare, the output component obtained from the software in time domain was compared
with the in-situ measurements (accelerograph) of station (2). The comparison between the
different values showed that the acceleration values in the time interval has increased 10 to 40
seconds and the highest acceleration has occurred on the time of 15th second which was about

7
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

0.205 g. By comparing the results obtained from the software and the results of the
accelerographs, we can carefully realize the results of the software.

The North-South Component of Earthquake

Figure 4: The time history of acceleration for north-south component obtained in station 2 (In-
situ measurements and calculated values)

In figure 4, the direct analysis was applied for north-south component. For comparison, we
compared the output component obtained in time domain with in-situ measurement
(accelerographs) of station (2). The comparison between the different values showed that the
acceleration values in the time interval has increased 10 to 30 seconds and the highest
acceleration has occurred on the time of 15th second which was about 0.242 g. By comparing
the results obtained from the software and the results of the accelerographs, we can carefully
realize the results of the software.

The Vertical Component of Earthquake

8
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 5: The time history of acceleration for vertical component obtained in station 2 (In-situ
measurements and calculated values)

In figure 5, the direct analysis was applied for vertical component. For comparison, we
compared the output component obtained in time domain with in-situ measurements
(accelerographs) of station (2). The comparison between different values demonstrated that
the acceleration values at the interval of 5 to 7 seconds show little difference in the results of
software and in-situ measurements values. The acceleration values in the time interval has
increased 10 to 35 seconds and the highest acceleration has occurred on the time of 21 th
second which was about 0.165 g. By comparing the results obtained from the software and the
results of the accelerographs, we can carefully realize the results of the software (except the
interval of 5 to 7 seconds).

3.2. Deconvolution Analysis in Time Domain

For deconvolution analysis, the input motion in station (2) is given to the software and the
output in station (1) is obtained from the software , and for controlling the accuracy, the
results obtained from the software are compared with the results obtained from
accelerographs in time domain.

9
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 6: A view of deconvolution analysis

The East-west Component of Earthquake

Figure 7: The time history of acceleration for east-west component obtained in station 1 (In-situ
measurements and calculated values)

As shown in figure (7), the east-west input motion of station (2) in ground surface is given
to the software and the output motion in station (1) in the surface of outcrop is obtained from
the software or in other words, the deconvolution analysis has taken place. For comparison,
we compared the output component obtained in time domain with in-situ measurements
(accelerographs) of station (1). The comparison between the different values showed that the
acceleration values in the interval of 5 to 30 seconds have increased and the highest
acceleration has occurred on 15th second which was about –0.178 g. The comparison between

10
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

the results obtained from the software and the results of the accelerographs showed that there
is a difference between the values obtained from the software and the results of
accelerographs.

The North-South Component of Earthquake

Figure 8: The time history of acceleration for north-south component obtained in station 1 (In-
situ measurements and calculated values)

In figure (2), deconvolution analysis is applied for north-south component. The comparison
between the different values showed that the acceleration values in the interval of 10 to 30
seconds have increased and the highest acceleration has occurred on 15th second which was
about –0.212 g. By comparing the results obtained from the software and the results of the
accelerographs, we can carefully realize the results of the software.

The Vertical Component of Earthquake

11
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 9: The time history of acceleration for vertical component obtained in station 1 (In-situ
measurements and calculated values)

In figure (9), such as the previous components, we have used deconvolution analysis for
vertical component. The comparison between the different values showed that the
acceleration values in the interval of 7 to 40 seconds have increased and the highest
acceleration has occurred on 21th second which was about 0.135 g. The comparison between
the results of software and accelerographs’ results show a little difference among the results.

3.3. Direct Analysis in Frequency Domain

The Fourier amplitude spectrum for strong ground motion shows the range distribution by
frequency (or period) and it clearly draws the frequency content of the motion. The Fourier
amplitude spectrum can be narrow or wide, where the narrow range shows that the motion has
a dominant frequency which can produce an almost flat time history (accelerograph), while on
the other hand, a broad range has various frequencies which produce an irregular and uneven
time history (accelerograph). Similar to the previous case, for direct analysis the results of
time history of given acceleration is considered as input motion at the surface of outcrop in
station (1) and the output in station (2) placed on the ground surface is obtained from the
software, and then the results obtained from the software is compared with the results
obtained from accelerographs in time domain in order to ensure the accuracy of the results.
All the steps above are done for east-west, north-south, and vertical components and the
results are as follows.

12
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 10: A view of direct analysis

The East-west Component of Earthquake

Figure 11: The Fourier amplitude spectrum for east-west component obtained in station 2 (In-
situ measurements and calculated values)

As it’s demonstrated in figure (11), the east-west input motion of station (1) in the surface
of outcrop is given to the software and the output motion of station (2) in the ground surface

13
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

is obtained from the software or in the other words, the direct analysis has taken place. In
order to compare, the output component obtained from the software in frequency domain was
compared with the in-situ measurements (accelerograph) of station (2). The comparison
between the different values in the Fourier amplitude spectrum showed that the amplitude
values are up to 15 Hz frequency, and the highest amplitude for the values obtained from
software occurs at frequency of 2.3 Hz and for in-situ measurement values, it is occurred at
frequency of 2.7 Hz. The obtained results from the software show a high accordance with the
results obtained from in-situ measurements.

The North-South Component of Earthquake

Figure 12: The Fourier amplitude spectrum for north-south component obtained in station 2
(In-situ measurements and calculated values)

In figure (12), the direct analysis was applied for north-south component. For comparison,
we compared the output component obtained in frequency domain with in-situ measurement
(accelerographs) of station (2). The comparison between the different values in the Fourier
amplitude spectrum showed that the amplitude values are up to 15 Hz frequency, and the
highest amplitude for the values obtained from software occurs at frequency of 3.02 Hz and
for in-situ measurement values, it is occurred at frequency of 3.03 Hz. The obtained results
from the software show a high accordance with the results obtained from in-situ
measurements.

The Vertical Component of Earthquake

14
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 13: The Fourier amplitude spectrum for vertical component obtained in station 2 (In-situ
measurements and calculated values)

In figure (13), the direct analysis was applied for vertical component. For comparison, we
compared the output component obtained in frequency domain with in-situ measurement
(accelerographs) of station (2). The comparison between the different values in the Fourier
amplitude spectrum showed that the amplitude values are up to 15 Hz frequency, and the
highest amplitude for the values obtained from software occurs at frequency of 2.41 Hz and
for in-situ measurement values, it is occurred at frequency of 2.8 Hz. The obtained results
from the software show a high accordance with the results obtained from in-situ
measurements.

3.4. Deconvolution Analysis in Frequency Domain

For deconvolution analysis, the input motion in station (2) is given to the software and the
output in station (1) is obtained from the software, and for controlling the accuracy, the results
obtained from the software are compared with the results obtained from accelerographs in
frequency domain.

15
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 14: A view of deconvolution analysis

The East-west Component of Earthquake

Figure 15: The Fourier amplitude spectrum for east-west component obtained in station 1 (In-
situ measurements and calculated values)

As shown in figure (15), the east-west input motion of station (2) in ground surface is given
to the software and the output motion in station (1) in the surface of outcrop is obtained from
the software or in other words, the deconvolution analysis has taken place. For comparison,
we compared the output component obtained in frequency domain with in-situ measurements
(accelerographs) of station (1). The comparison between the different values in the Fourier
amplitude spectrum showed that the amplitude values are up to 15 Hz frequency, and the

16
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

highest amplitude for the values obtained from software occurs at frequency of 2.98 Hz and
for in-situ measurement values, it is occurred at frequency of 2.29 Hz. For the frequencies of
3 to 5 Hz, there is a little difference between the results obtained from the software and the
results obtained from accelerographs.

The North-South Component of Earthquake

Figure 16: The Fourier amplitude spectrum for north-south component obtained in station 1
(In-situ measurements and calculated values)

In figure (16), deconvolution analysis is applied for north-south component. The

comparison between the different values in the Fourier amplitude spectrum showed that the
amplitude values are up to 17 Hz frequency, and the highest amplitude for the values obtained
from software occurs at frequency of 1.52 Hz and for in-situ measurement values, it is
occurred at frequency of 4.56 Hz. For the frequencies of 3 to 8 Hz, there is a little difference
between the results obtained from the software and the results obtained from accelerographs.

The Vertical Component of Earthquake

17
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 17: The Fourier amplitude spectrum for vertical component obtained in station 1 (In-situ
measurements and calculated values)

In figure (17), deconvolution analysis is applied for north-south component. The

comparison between the different values in the Fourier amplitude spectrum showed that the
amplitude values are up to 20 Hz frequency, and the highest amplitude for the values obtained
from software occurs at frequency of 2.88 Hz and for in-situ measurement values, it is
occurred at frequency of 2.41 Hz. For frequencies of 3 to 8 Hz and 10 to 25 Hertz, there is a
little difference between the results obtained from the software and the results obtained from
accelerographs.

4. Comparing the Obtained Results in Time and Frequency Domains

In the last sections of this paper, the results obtained from DEEPSOIL software in time and
frequency domains were compared with the values obtained from in-situ measurements
(obtained from accelerographs). The comparison between the results of software and
accelerographs demonstrated that the software results have more accuracy; therefore, after
verify the accuracy of software results, in this section, we compare the results obtained for
east-west, north-south and vertical stations in both time and frequency domains and the results
are as follows.

18
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 18: A view of direct and deconvolution analysis

The East-west Component of Earthquake

Figure 19: The acceleration time history of the east-west component of earthquake obtained
from station 1 and 2

As shown in figure (19), In order to compare, the east-west component has been compared
in direct and deconvolution analysis. The results from direct analysis and in time domain, i.e.
from station (1) to station (2), showed that the acceleration values increased after passing the
soil layers and we observed acceleration enlargement in the ground surface. This phenomenon
is more obvious in the interval of 5 to 30 seconds. The results from deconvolution analysis
and in time domain, i.e. from station (2) to station (1), showed that the acceleration values
decreased after passing the soil layers and reaching the seismic bedrock (the maximum
acceleration in the time of 15 seconds has decreased from .021 g to 0.11 g). The decrease in
acceleration values is more obvious in the interval of 10 to 20 seconds. While the acceleration
values in the interval of 30 to 50 seconds are almost constant.

19
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 20: The Fourier amplitude spectrum of the east-west component of earthquake obtained
from station 1 and 2

As we mentioned earlier, the results are compared in the frequency domain as well as time
domain. As it is obvious in figure (20), the results obtained from the direct analysis of east-
west input motion show that the according Fourier spectrum value of frequency of 2.3 at the
surface of outcrop has increased from 0.11 to 0.25 in the ground surface, which indicates the
occurrence of amplitude phenomenon in the ground surface or in the other words, the
earthquake waves have experienced amplification phenomenon in ground surface after
passing the soil layers, this phenomenon (amplification) is more evident at frequencies of 3 to
10 Hz. The results of deconvolution analysis showed that the Fourier spectrum value of
frequency of 2.3 at the ground surface has decreased from 0.25 to 0.11 in the surface of
outcrop, which indicates the amplification decrease at this frequency. The decreased
amplification values are more obvious at frequencies of 3 to 10 Hz.

The North-South Component of Earthquake

20
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 21: The acceleration time history of the north-south component of earthquake obtained
from station 1 and 2

As shown in figure (21), In order to compare, the north-south component has been
compared in direct and deconvolution analysis. The results from direct analysis and in time
domain, i.e. from station (1) to station (2), showed that the acceleration values increased after
passing the soil layers and we observed acceleration enlargement in the ground surface. This
phenomenon is more obvious in the interval of 12 to 50 seconds. The results from
deconvolution analysis and in time domain, i.e. from station (2) to station (1), showed that the
acceleration values decreased after passing the soil layers and reaching the seismic bedrock
(the maximum acceleration in the time of 14.68 seconds has decreased from .02 g to 0.103 g).
The decrease in acceleration values is more obvious in the interval of 10 to 20 seconds. While
the acceleration values in the interval of 12 to 30 seconds are almost constant.

Figure 22: The Fourier amplitude spectrum of the north-south component of earthquake
obtained from station 1 and 2

21
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

As it is obvious in figure (22), the results obtained from the direct analysis of north-south
input motion show that the according Fourier spectrum value of frequency of 3.027 at the
surface of outcrop has increased from 0.055 to 0.203 in the ground surface, which indicates
the occurrence of amplitude phenomenon in the ground surface or in the other words, the
earthquake waves have experienced amplification phenomenon in ground surface after
passing the soil layers, this phenomenon (amplification) is more evident at frequencies of 3 to
8 Hz. While the amplification values at 6 to 9 Hz have decreased to some extent. The results
of deconvolution analysis showed that the Fourier spectrum value of frequency of 3.027 at the
ground surface has decreased from 0.203 to 0.055 in the surface of outcrop, which indicates
the amplification decrease at this frequency. The decreased amplification values are more
obvious at frequencies of 3 to 8 Hz. While the amplification values at 6 to 9 Hz have
increased to some extent.

The Vertical Component of Earthquake

Figure 23: The acceleration time history of the vertical component of earthquake obtained from
station 1 and 2

As shown in figure (23), the results from direct analysis and in time domain, i.e. from
station (1) to station (2), showed that the acceleration values increased after passing the soil
layers and we observed acceleration enlargement in the ground surface. This phenomenon is
more obvious in the interval of 7 to 32 seconds. While the acceleration values have decreased
in the interval of 5 to 8 seconds. The results from deconvolution analysis and in time domain,
i.e. from station (2) to station (1), showed that the acceleration values decreased after passing
the soil layers and reaching the seismic bedrock (the maximum acceleration in the time of
21.39 seconds has decreased from 0.16 g to 0.07 g). The decrease in acceleration values is
more obvious in the interval of 7 to 32 seconds.

22
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

Figure 24: The Fourier amplitude spectrum of the vertical component of earthquake obtained
from station 1 and 2

As it is obvious in figure (24), the results obtained from the direct analysis of vertical input
motion show that the according Fourier spectrum value of frequency of 2.41 at the surface of
outcrop has increased from 0.06 to 0.18 in the ground surface, which indicates the occurrence
of amplitude phenomenon in the ground surface or in the other words, the earthquake waves
have experienced amplification phenomenon in ground surface after passing the soil layers,
this phenomenon (amplification) is more evident at frequencies of 3 to 10 Hz. The results of
deconvolution analysis showed that the Fourier spectrum value of frequency of 2.41 at the
ground surface has decreased from 0.18 to 0.06 in the surface of outcrop, which indicates the
amplification decrease at this frequency. The decreased amplification values are more obvious
at frequencies of 3 to 10 Hz.

Conclusion

In this paper, the components obtained from seismograph devices, which recorded 6
components for the project, were first converted to time history for acceleration. These
components included east-west, north-south and vertical components. The field values (in-situ
measurements) were modified and filtered which was done with band-path filter (of 0.2-30
Hz) and modifying the baseline. After modifying the accelerographs, the analysis was
performed in direct and deconvolution ways and the results were presented in the form of
time history acceleration and the Fourier amplitude spectrum. In order to verify the accuracy
of the results obtained from the DEEPSOIL software, the results were compared with the field
values (in-situ measurements) in both time and frequency domains for all the components,

23
 International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

which indicated a higher accuracy in software results. The results obtained from these
analyses are as follows:
1) The earthquake waves are strengthened or weakened at specific frequencies after releasing
from seismic bedrock and moving toward ground surface due to their dynamic characteristics
of the alluvium. If this frequency equals with the structure’s frequency, the amplification
phenomenon will take place which leads to structure destruction.
2) From an earthquake engineering point of view, the most important properties of strong
ground motion include domain, frequency content and duration. All these properties have a
great impact on earthquake destruction.
3) The fact that the earthquake has broader destructions in some areas and some minor
destruction in other places represents the importance of local site effects.
4) An earthquake may be strengthened or weakened at certain frequencies due to the soil
properties. The results obtained from direct analysis in time domain for east-west, north-south
and vertical components showed that the acceleration values for waves increase after passing
the alluvial layers (soil layers). The results obtained from deconvolution analysis showed that
the acceleration values for waves decreased after passing soil layers and in the way of
reaching the seismic bedrock.
5) The results obtained from direct analysis in frequency domain for east-west, north-south
and vertical components showed that the earthquake waves encounter with amplification
phenomenon at the ground surface after passing the alluvial layers (soil layers). The results of
deconvolution analysis also showed that amplification values has decreased at some certain
frequencies.

References
[1] Soghrat M.R et al, Evaluation of Site Effects of North of Iran By Using Empirical method H/V,
The Sixth National Congress on Civil Engineering, Semnan University, Semnan, 2011.
[2] Kramer, Steven L., Geotechnical Earthquake Engineering, 1st ed., New Jersey, Prentice-Hall, 1996.
[3] Joyner, W.B. and Chen, Calculationof Nonlinear Ground Response in Earthquake, bulletin of the
seismological society of America, 1975.
[4] Hwan, G and Lee, Parametric Study of Sitr Response Analysis, Soil dynamic and Earthquake
Engineering, 1991.
[5] Mir Hosseini, M.A and Asadolahi Pajoh, M, The study of Site Effects on Strong Ground Motion in
Alluvial Areas Adjacent Bandare Mahshahr, The Eighth International Congress on Civil Engineering,
Shiraz University, Shiraz, 2009.
[6] Hashash, Y.M.A, DEEPSOIL V3.7, User Manual and Tutorial, Department of Civil and
Environmental Engineerin, University of Illinois at Urban-Champaign, 2009.

24
International Conference on Civil Engineering
Architecture & Urban Sustainable Development
18 &19 December 2013, Tabriz , Iran

25